INDEX

Explanations

references to the concept of lacking or the absence of various elements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 iload

-0.51

WARD

-0.50

warded

-0.48

IndentedString

-0.48

̀i

-0.48

 mxArray

-0.47

 AssemblyTitle

-0.46

原始内容

-0.45

mutations

-0.45

TagHelpers

-0.45

POSITIVE LOGITS

 else

1.88

 ELSE

1.12

 Else

1.10

Else

1.09

else

0.95

ELSE

0.88

 resembling

0.83

 besides

0.77

 except

0.69

besides

0.68

Activations Density 0.101%

references to the concept of lacking or the absence of various elements

No Comments

No Known Activations